公開日:
Work Analysis
OWAS vs. REBA: Which Method Should You Use to Assess Work Posture?

Work element: A worker leaning forward, raising arms while securing a connection, and retrieving parts from a low box. Even with a desire to reduce strain, it's difficult to determine which process to re-evaluate.
Work element: OWAS method and REBA method are techniques for observing work posture to assess strain and prioritize improvements.Work element: OWAS method organizes postures across the entire operation into broad categories, while REBA method details specific evaluation including neck and wrist.Work element: Points represent the primary distinction.
Work element: This article explains both methods' evaluation content and application, using the Ollo Factory unit assembly as an example. The focus is on organizing a flow that links desk and component layout improvements, not on assigning scores.
- Work element: Comparison of OWAS method and REBA method.
- Characteristic: Ability to record posture in relatively large categories - Characteristic: Ability to record posture in relatively large categories
- REBA method: confirm neck, wrist, and load-carrying posture in selected scenarios.
- Choose the method based on the work posture and scope
- 3 steps from evaluation to improvement
- 1. Gather a full-body video and working conditions.
- 2. Leave the evaluated posture and rationale.
- 3. Change configuration or method and re-evaluate under the same conditions.
- Frequently asked questions
- If REBA is more detailed, is OWAS unnecessary?
- Can a low score be considered safe?
- Can AI evaluate from a video?
- Summary: Use based on the scope and area to be evaluated.
- "Ollo Factory", a manufacturing-specialized work analysis AI that automates the evaluation of work posture.
- Watch Ollo Factory's work posture evaluation AI in video.
- Watch Ollo Factory’s Ergonomic Assessment AI in Action
Work element: Comparison of OWAS method and REBA method.
Work element: Both techniques handle the entire body posture, but the division of body parts and evaluation conditions differ.
Work element: Comparison criteria. | Work element: OWAS method. | Work element: REBA method. |
|---|---|---|
Work element: Main focus body parts. | Work element: Back, limbs, and lower body. | Work element: Torso, neck, lower body, and upper arm, forearm, and wrist. |
Work element: Conditions other than posture. | Work element: Load and force handled. | Work element: Load, force, ease of handling, posture maintenance, and repetitive activities. |
Work element: Recording approach. | Work element: Code posture. Observe at regular intervals and organize appearance patterns. | Work element: Review the entire operation to select target situations and evaluate each body part's angle and conditions. |
Work element: Method of expressing results. | Four action categories (AC1–AC4) | Score-based 5-level response classification - Organize into 5 levels based on score |
Easily applicable scenarios - Usable scenarios | Understand posture distribution within the process to pinpoint improvement targets - Determine improvement targets by understanding posture distribution within the process | Conduct detailed evaluation of concerning work scenarios and compare before and after - Conduct detailed evaluation of concerning work scenarios and compare before and after |
OWAS method overview - OWAS method overview is provided by the Agricultural Research InstitutePublic materials on posture assessment - Public materials on posture assessment are also introducedOWAS method and REBA method evaluation criteria - OWAS method and REBA method evaluation criteria are found in the 2nd chapter of the University of Tokyo's public research materialsCannot directly compare scores of both methods - Scores of both methods cannot be directly comparedFor instance, OWAS AC2 and REBA score 2 do not imply the same workload or response - For instance, OWAS AC2 and REBA score 2 do not mean the same workload or response; comparisons are made before and after using the same method and conditions
OWAS method: Identify concerning posture across the entire process - OWAS method: Identify concerning posture across the entire process In OWAS method, classify back curvature, twisting, arm height, leg condition, and load, then match with evaluation table and organize into action categories (AC) - In OWAS method, classify back curvature, twisting, arm height, leg condition, and load, then match with evaluation table and organize into action categories (AC)
Characteristic: Ability to record posture in relatively large categories - Characteristic: Ability to record posture in relatively large categories
For example, in unit assembly: Observing posture at regular intervals reveals which tasks exhibit forward-leaning posture and its frequency. Maintain observation time, task name, and posture code to easily return to the improvement scenario.
For example, in unit assembly: Detailed distinction for neck and wrist angle is not applicable. Even tasks appearing in the same OWAS category may differ in tool handling and wrist flexion.
For example, in unit assembly: Detailed distinction for neck and wrist angle is not applicable. Even tasks appearing in the same OWAS category may differ in tool handling and wrist flexion.
Unit assembly example: Observation during component supply, assembly, fastening, inspection reveals, "bending back every time I take parts from the lower box," indicating a work element for improvement. Short-term heavy load may occur during observation intervals, so supplement with operator complaints and video review.
Posture code recording method is detailed in.OWAS method evaluation procedure.Detailed explanation provided.
REBA method: confirm neck, wrist, and load-carrying posture in selected scenarios.
REBA method is an acronym for Rapid Entire Body Assessment, a technique to evaluate full-body work posture. Separate recording of torso, neck, lower limbs, upper arm, forearm, and wrist, combining load, force, ease of handling, and activity conditions.
After observing the work once, select frequent postures, complaint-based postures, and postures noted during surveillance to clearly define the work element for improvement, not to assign the highest score.
Unit assembly example: evaluate the situation when fastening the back of the fixture by bending the neck, raising the elbow, and twisting the wrist. Record tool weight and ease of handling; re-evaluate after changing the fixture orientation and tool support method. It is important not to synthesize individual postures into one. Scoring the bending of the waist when taking parts and the arm height during fastening as a single posture may result in evaluating a scenario that does not actually exist.
Work posture-specific differentiation.
Choose the method based on the work posture and scope
Choose a method by deciding "which part you want to see" and "whether you want to examine the entire process or focus on a specific scene in detail".
What to confirm in the field | Example of proceeding | Additional information to retain |
|---|---|---|
Where in multiple processes is the posture burden most frequent? | Use OWAS to observe the entire process and organize the appearance of posture | Process name, observation time, observation count, frequency of work |
Does the inner fastening create a burden on the neck or wrists? | Use REBA to evaluate the target fastening scene in detail | At the moment of evaluation, the target side, tools/load, posture maintenance and repetition conditions |
Do you want to confirm the effect of changing the component box or fixture? | Re-implement the method used before under the same working conditions | Box height, fixture orientation, variety, worker, evaluation conditions |
You do not necessarily have to use both. You can use REBA to help narrow down candidates with OWAS and to gain a detailed understanding of the state of the neck and wrists in situations where you want to know. Choose the method that provides the necessary information for your initial objective.
3 steps from evaluation to improvement
1. Gather a full-body video and working conditions.
Shoot from a position where the head, torso, arms, legs, and feet can be seen. A video where only the hands are visible cannot determine the full-body posture. If the body is hidden by fixtures or equipment, add a different angle or on-site confirmation.
Define the target: Specify variety, process, action to be evaluated, and distinguish normal work from exceptions.
Record conditions: Retain the weight of parts/tools, height of work table, number of repetitions, etc.
Verify visibility: Confirm whether the feet and arms are out of frame, and whether left/right and depth can be read.
Place cameras and tripods in positions that do not obstruct the workflow. Do not force the operator into a different posture than usual for shooting; record according to the field shooting rules.
2. Leave the evaluated posture and rationale.
OWAS records observation method and posture code; REBA records target scene and evaluation conditions per body part. Do not estimate unseen body parts or unknown weights by guesswork; classify as items requiring further verification.
Ensure the video timestamps and evaluation tables are synchronized so that the process can be reviewed even if the operator changes. When results diverge, verify posture reading and load conditions before averaging scores.
3. Change configuration or method and re-evaluate under the same conditions.
For example, if retrieving low-level parts boxes is burdensome, consider lifting the box or moving the supply position closer. If rear fastening is difficult, re-examine fixture orientation, work position, and tool support method.
After improvement, assess posture evaluation, work time, quality missed checks, and additional body part burden. Even if burdensome postures decrease, check for narrowing workflow or increased workload.
To reduce effort in posture confirmation and recording, video-based AI analysis may also be a benchmark. The corresponding evaluation methods and shooting conditions can be confirmed in the product documentation.
Frequently asked questions
If REBA is more detailed, is OWAS unnecessary?
It is not necessary. If the objective is to continuously record the posture of the entire process, the OWAS method may be suitable as it is easy to use, especially when not choosing based on detail alone but rather on the evaluation target and the labor time required for operation.
Can a low score be considered safe?
The posture evaluation results alone cannot guarantee the safety of the entire operation. Equipment entanglement or falls are other hazards. Additionally, the body type, physical condition, and burden of prolonged work must be verified and incorporated into the site's safety and hygiene management.
Can AI evaluate from a video?
There is a method to estimate the skeleton and joint angles to assist with evaluation. It is necessary to confirm the overall appearance and load from the video, including conditions not visible in the image. The AI results and the supporting video are compared for use.
Summary: Use based on the scope and area to be evaluated.
The OWAS method is easily applicable for organizing the posture of the entire process and narrowing down the improvement targets.
The REBA method can evaluate specific situations, including the neck, wrists, and ease of handling.
Compare before and after using the same method and conditions to verify quality and other burdens.
Automating bone detection from videos and creating evaluation tables with AI reduces the burden of observation and recording, making it easier to allocate time for reviewing improvement proposals and confirming effectiveness.
"Ollo Factory", a manufacturing-specialized work analysis AI that automates the evaluation of work posture.
Ollo Factory analyzes the skeleton from full-body work videos to support posture evaluation using OWAS method and REBA method. You can check evaluation for each work element and time series load, and re-examine results by changing conditions such as load.

Public demo screen of Ollo Factory official website. Not a result of analyzing the process for text description.
For sites where it is burdensome to read and record posture one by one in a table, please consult regarding application in actual videos after organizing shooting conditions and evaluation targets. We will guide you on how to proceed from posture evaluation to improvement consideration.
Watch Ollo Factory's work posture evaluation AI in video.
Watch Ollo Factory’s Ergonomic Assessment AI in Action
この記事の執筆者
株式会社Ollo
事業開発部リーダー マーケティング担当 Yoshida
製造業特化型の作業分析AI「Ollo Factory」を開発・提供する株式会社Ollo(東京大学 松尾研究室発のAIスタートアップ)で、事業開発とマーケティングを担当。
大学時代は工学部で設計開発を学び、新卒で大手自動車メーカーに入社。新型EVの開発グループで、設計から工場での生産まで幅広くプロジェクトに携わり、研修の一環として、実際に自動車製造工場でのライン作業も経験。その後、製造業領域のスタートアップで営業・マーケティングに従事。
現在は、製造現場でのAI活用を促進し、よりワクワクする働き方を実現できるよう、「Ollo Factory」の販売・マーケティングに取り組んでいる。
一覧に戻る
関連記事
Ollo Factoryについて、さらに詳しく。

まずは資料請求
・ 導入事例と効果
・ 主な機能
・ 活用の流れ

